1. Field of the Invention
This invention relates to a welding method for use in building large-scale structures such as ships, bridges and the like. More particularly, the invention relates to a welding method which is highly capable of imparting improved fatigue strength to weld joints in the structures. The invention further relates to special welding materials such as welding rod or welding wire and the like for the practice of the method.
2. Description of the Rpeated Art
In recent years ships, marine structures, penstocks and the like have increased significantly in size. A demand has arisen for steel members having improved strength in order to attain weight saving. So-called low-alloy ferrous steels have hitherto been used that contain alloy elements such as Cr, Ni, Mo and the like in amounts of less than 3.0 wt %. Such steel members have a tensile strength of 30 to 120 kgf/mm.sup.2.
To cope with the demand for strength improvement, steel members of greater strength may be selected from various low-alloy ferrous steels. The fatigue strength of such a high-strength low-alloy ferrous steel increases as the strength of the corresponding base metal increases. However, it is generally known that a weld joint derived from such low-alloy steel fails to attain improved fatigue strength even when the associated strength of weld joint is increased (please see Lecture Resumes of National Meeting, Welding Society of Japan, No. 52, Pages 256-257, 1993).
Consequently, a weld joint obtained from a high-strength steel shows the same level of fatigue strength as that of steel of a low strength. This presents the problem that, in the case of those structures calling for weld joints as by fillet welding, designed strength cannot be increased as desired, even with the use of a steel member of high strength. Excessively high residual tensile stress in the weld results in failure to achieve improved fatigue strength of the weld joint, even one resulting from use of such a high-strength steel member (please see Papers, Welding Society of Japan, Vol. 13, No. 3, Pages 438-443, 1995).
Another reason for the inability to improve fatigue strength is the fact that a weld joint derived from high-strength steel has great notch sensitivity. This problem, however, can be solved by using an improved welding process, or by increasing the radius of a weld toe through smooth grinding, or by use of a rotary cutter or the like as disclosed in Japanese Unexamined Patent Publication No. 5-69128.
Residual tensile stress occurs in the weld zone of the joint. One reason for this is that a weld metal formed in the welding contracts during subsequent cooling. In FIG. 18 of the accompanying drawings, the weld metal is contracted in the course of cooling upon completion of welding of a low-alloy steel member using a conventional welding material formed from a low-alloy steel. This weld metal, after welding, causes thermal contraction in the direction indicated by the arrow in FIG. 18 of the drawings.
In such instance, when use is made of a welding material derived from conventional low-alloy steel, the weld metal thermally contracts and gradually declines in elongation (length) with temperature drop, but expands or increases in elongation (length) in the neighborhood of 500.degree. C. This is because martensite transformation occurs at about 500.degree. C. and subjects the weld metal to expansion. Upon completion of the martensite transformation, the weld metal again undergoes thermal contraction only and undergoes reduced elongation as the temperature declines. For its low yield stress, the weld metal is cooled under plastic deformation during cooling from the point of solidification to about 600.degree. C. Plastic deformation serves to relax residual tensile stress arising from contraction of the weld metal. In the case of contraction at a temperature below 600.degree. C., the weld metal suffers from high yield stress, thus becoming less likely to undergo plastic deformation, with ultimate introduction of residual tensile stress.
As is shown in FIG. 18, the weld metal cooling from 900.degree. C. or 1000.degree. C. contracts until it reaches a temperature of about 500.degree. C. to 600.degree. C. but then expands by virtue of martensite transformation in the course of cooling from about 500.degree. C. to about 400.degree. C., during which the residual tensile stress is relaxed. However, in the course of cooling to room temperature after martensite transformation thermal contraction takes place during the subsequent temperature drop to room temperature, and the resulting weld metal develops considerable residual tensile stress.
This explanation shows the chief cause for weld joint involvement in residual tensile stress. Where two different steel members to be welded have different sizes, for example as in fillet welding, any residual tensile stress is amplified due to the difference in heat capacity between the two steel members.
When a T-shaped joint is formed by fillet welding (as seen in FIG. 19A), a main plate 1 and a sub-plate 2 are welded in a lengthwise direction 4 and in the direction of arrow A as shown in FIG. 19B. During this welding, the temperatures of the main plate 1 and of the sub-plate 2 are gradually raised through the heat applied to the zone to be welded. When the sizes of the main plate 1 and of the sub-plate 2 are different from each other, the magnitudes of thermal diffusion vary as between the plates. When it has a smaller volume than that of the main plate 1, the sub-plate 2 experiences a larger temperature rise. Hence the amount of thermal expansion produced upon welding is greater in the sub-plate 2 than in the main plate 1. This leads to a difference in thermal expansion between the main plate 1 and the sub-plate 2 while welding the weld zone (1). This difference affects subsequent welding of a subsequent weld zone (2). Since the weld zone (2) is welded under such a difference in thermal expansion, a residual tensile stress takes place in the weld zone (2) during thermal contraction in cooling after welding.
Furthermore, where a joint is formed by bringing a sub-plate 2 into welded relation to a main plate 1 by swivel fillet welding as illustrated in FIGS. 20A and 20B, the sub-plate 2 may be swiveled in the order of a long side ((1) of FIG. 21)--a short side ((2) of FIG. 21)--((3) of FIG. 21) viewed from a cross section 3 of a weld zone thereof. According to the procedure of FIG. 21, welding is effected in two cycles of swiveling since the weld zone 3 of the sub-plate 2 is rather large in size. Also in this instance, the sub-plate 2 has a greater temperature rise because of its smaller volume than that of the main plate 1, and hence, the amount of lengthwise elongation is greater in the sub-plate 2 upon welding of the long side. Residual tensile stress thus takes place in the weld zone in the course of thermal contraction cooling after welding.
When the residual tensile stress resulting from the difference in thermal expansion between the main plate and the sub-plate is brought about concurrently with residual tensile stress arising from thermal contraction of the weld metal, as discussed previously, the resulting residual tensile stress often rises almost to the yield strength of the finished weld joint.
In order to reduce the residual tensile stress in a weld joint, one method has been disclosed in Japanese Unexamined Patent Publication No. 4-21717. A sub-plate is joined to a main plate by swivel fillet welding, followed by hammering the welding toe as by a shot peening, hammer peening or the like, thereby imparting compression stress to such toe and hence reducing the residual tensile stress in the welding toe. In this method, however, special after-weld treatment needs to be conducted after welding. Since machinery and operation not commonly used for welding are required, this prior art method can be said to be neither effective nor economical.
To reduce residual tensile stress in a weld zone, Japanese Unexamined Patent Publication No. 54-130451 discusses a method wherein at least the final or outermost layer of a weld metal is welded with a welding material composed of an austenite type iron alloy, whereby martensite transformation is allowed to initiate at a temperature below room temperature, and the resulting weld is then cooled with use of liquid nitrogen, for example, at a temperature lower than -60.degree. C. Also in the same method, cooling at a temperature of lower than -60.degree. C. with liquid nitrogen is needed after common welding. In use of steel plates of especially large thickness, this is excessively tedious, expensive and time-consuming because it is necessary to cool all of the weld zone at -60.degree. C. or below with liquid nitrogen.